Prognostic Factor Analysis Including Electrocardiogram Change in Patientss with Subarachnoid Hemorrhage

ABSTRACT

PURPOSE: The prognostic factors of subarachnoid hemorrhage (SAH) are still not completely known. Several studies suggested that electrocardiogram (ECG) changes can act as a predictor of outcome in SAH patients. The purpose of this study was to describe the prognostic factors, including ECG changes, which are predictive of unfavorable outcome in non-traumatic SAH patients. METHODS: We retrospectively selected patients from our prospectively collected database of 202 SAH patients who visited the emergency medical center. The outcome was assessed using the Glasgow Coma Scale at six months after the occurrence of SAH. RESULTS: In the univariate analysis, a high score in one of the conventional systems (Hunt and Hess system, World Federation of Neurosurgical Societies [WFNS] scale, and Fisher grade), advanced age, accompanying intracranial hemorrhage or intraventricular hemorrhage, ECG changes (ST depression or Tall T), and a history of hypertension were associated with unfavorable outcome. The multivariate analysis showed three prognostic factors (ECG changes, age and high score in the conventional system) for unfavorable outcome. Using this result, three novel models corresponding to the three conventional systems were constructed to predict an unfavorable outcome in such patients. The area under the curve for model 1 (containing the WFNS scale) was 0.912, that of model 2 (containing the HH system) was 0.913, and that of model 3 (containing the Fisher system) was 0.885. Compared with the WFNS, HH or Fisher grade alone, each model exhibited superior accuracy. CONCLUSION: ECG can be described as an independent predictor of poor outcome, and the novel models which contain the ECG changes were found to be more accurate in predicting an unfavorable outcome in SAH patients compared with the conventional scoring system.